DDraceNetwork - developer

fokkonaut

[ddnet7:master] 1 new commit

09ea52c Fix mute - fokkonaut

ddnet-maps

[ddnet-maps:master] 1 new commit

84aabfc M Redefined, M Dummy Chamber - ddnet-maps

@heinrich5991 no clue how to debug such thing. Any ideas?

watch/breakpoints in the graphics thread

and what am i searching there?

@ChillerDragon try gfx_resizable 0

Yea that helped thanks @Deleted User I will play around the next time ill need to launch the client but for now it sometimes worked.

@heinrich5991 nice blogpost

where did u read it?

09:07

do youe men github/heinrich/blog ?

https://heinrich5991.github.io/blog/blog/one-tick-unfreeze

The Anatomy of a One Tick Unfreeze

I saw the deepfly bind the other day:

ahty

09:08

i even starred the repo and didnt notice it was published xd

@heinrich5991 anything new this patch allow that we might not want?

cool blog post love it. But yikes the patch sounds really game changing.

09:20

I suggest letting some pros play with the patch first before merging it. And see what becomes possible. So that he do not risk some unexpected op behaviour.

09:20

would this allow for freeze hammer chaining for example?

if hammer buffers like grenade then yes

could be op on some maps i guess

you could easily hammer through thick freeze walls

for sure a t0 speedrunning game changer

09:25

i mean idk maybe not too bad and i really like the idea of supporting it without deepfly bind but we have to test it for sure,

ye it is cool but probably causes some cheats on old maps

Hmm, you could do hammer chains with a scrollwheel

also completely removes the timing aspect of "double hammer" out of unfreeze during a hammerhit

10:03

definitely needs testing

on the other hand it was the same with weapons in freeze before deen fixed it way back in 2013. maybe this should have been added years ago as well

@heinrich5991 wouldnt that allow deepflying with normal players?

10:23

basically means cheating a huge chunk of our current deepfreeze parts

10:23

(which are already cheatable with dummy - which is bad - but not intended to be done that way)

maybe just allow it on dummy and new maps first

10:39

what if its possible to just walk infinitly through freeze tee_thinking

having inconsistent mechanics from map to map is rly bad xd

10:39

just on dummy would be better that what we have now tho ye

https://github.com/teeworlds/teeworlds/commit/7df4f9c8bfed13f9f0e6216cc93dc263ff85f8a6 might be worth cherrypickin

check map hash before embedding it to the demo file. closes #2229 ...

wait,0.7 has sha256 hashes?

11:09

would never think they'd implement that citing something like map names too long or ugly

sure 0.7 has it all

11:19

did i mention already that 0.7 might be worth taking a look at and updating to it? xd

@ChillerDragon did I mention a lot of people need to pitch in to get ddnet to 0.7?

hi

didn't we agree that we'd get a lite ddnet version running on 0.7?

11:28

isn't it only missing sql?

y

wouldn't implementing sql be better then asking me?

11:29

I'm not really interested in porting the broken sql code as is

upgrading sql good sounds good

I'm exploring web assembly now, maybe i'll look at rewriting sql tomorrow

would be amazing

11:31

(hope ur not trollin me rn :c)

@heinrich5991 is the stopper fixed now completely? if yes, can you commit it to ddnet7?

@heinrich5991 can you please be here to assist me with some cmake?

@redix have a minute to spare

About the sql stuff... You want to stick to the old db scheme I think? I started implementing a sqlite based backend (should be easy to add MySQL aswell) but I'm using IDs instead of names and right now it only implements finishes and no teams, scores etc.

@redix I was thinking of working on the db scheme aswell

12:14

What do you mean IDs tho? players don't really have ids when we don't have accounts

12:14

@redix if you have a couple seconds to spare can you try building this? https://github.com/wasmerio/wasmer-c-api

wasmerio/wasmer-c-api

Example of the C API to embed the Wasmer runtime. Contribute to wasmerio/wasmer-c-api development by creating an account on GitHub.

Table for players and primary key as id

Their cmake example can't possibly work (and it doesn't) but they wouldn't have commited it if it doesn't work

And in the finish you use the id instead of the name. So you don't have to store the name multiple times in the finish table

12:16

But it makes only sense in 0.7 since you cannot change the name

What a joke that change is

12:17

If I ever get a workable plugin system, i'll fork teeworlds and it'll be the first change I'm reverting

Haha :D well I'm not sure why you would change the name... But not being able the change the skin was really a bad idea

Those things belonged in the snap and the couple dozen bytes saved by the initial snapshot being a full snapshot is not worth it

Yes

or maybe if I'm feeling particularly nice I'll implement it as a new extended snapobject so vanilla players can still connect but see the wrong names

It just bloats the code since you have to handle it differently in demos

@redix did you try compiling the wasmer example?

12:20

"will you?" is a better question if not I'll bug someone else

I'm on a phone right now

ah

12:21

@archimede67 hello my fellow developer, would you mind compiling a small example for me on this fine day?

12:22

@fokkonaut or you?

what?

can you give https://github.com/wasmerio/wasmer-c-api a quick compile?

wasmerio/wasmer-c-api

Example of the C API to embed the Wasmer runtime. Contribute to wasmerio/wasmer-c-api development by creating an account on GitHub.

oof hpow

just clone

12:23

then cmake . && make

why dont you do it? (just wondering, no offense)

It doesn't work for me, I'm wondering whether they commited a broken example or it's my cmake version or it's another macOS oddity